【24h】

SCIMA: A Novel Processor Architecture for High Performance Computing

机译:SCIMA:用于高性能计算的新型处理器体系结构

获取原文

摘要

Clock acceleration and ILP extraction have drastical-ly improved processor performance. However, perfor-mance of memory has not been improved so much as that of processor, thus the problem of performance gap between memory and processor, which is called memory wall problem, is becoming very serious. This will get more serious in the near future. We propose a new VLSI architecture which can solve this problem. Our architecture integrates software-controllable memory into the processor chip. We have evaluated our architecture on Linpack benchmark and QCD simulation. These results indicate that our proces-sor achieves high performance by using On-Chip Mem-oory effectively.
机译:时钟加速和ILP提取大大提高了处理器性能。但是,存储器的性能没有像处理器那样得到很大提高,因此存储器和处理器之间的性能差距问题(称为内存墙问题)变得非常严重。这将在不久的将来变得更加严重。我们提出了一种新的VLSI架构,可以解决此问题。我们的架构将软件可控存储器集成到处理器芯片中。我们已经在Linpack基准测试和QCD仿真上评估了我们的体系结构。这些结果表明,我们的处理器通过有效地使用片上存储器实现了高性能。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号